While many children receive high quality health care, recent national reports indicate there are circumstances when this is not true. The Commonwealth Fund Report on quality of care for children from 2004 provides evidence that children do not always have accessible or timely health care.3 For example, children who are not insured for a full year have been found to be nearly twice as likely to miss out on preventive care visits, compared to children who are insured.4 To further examine these topics, this Chartbook covers these issues, as well as others that affect coverage, access, unmet needs, and quality of health care for children, on a state and sub-state level. The level and detail of information provided may help to reveal opportunities for improvement in child health and health care across Ohio.
Survey Methods
While many data sources characterizing child health and socioeconomic conditions exist, few provide a comprehensive picture of the health care children receive in the context of those characteristics. The Ohio Family Health Survey (OFHS), conducted most recently in 2008 and 2004, is the primary data source for this site.
The OFHS is a random digit dialing telephone survey. The OFHS population includes all non-institutionalized adults and children residing in a home in Ohio. The sampling frame included households in this population with a working landline or cell phone number. Over 13,000 children ages 17 and under were included in the sample for the OFHS.
In each household contacted, an adult respondent was identified and surveyed for health data (including height, weight, and health status) and demography. When the household included a child aged 17 years or younger, the adult self-identified as having the most knowledge about the child’s health was also asked a series of questions about the child’s height, weight, health status, and demography. Responses for children with unknown and refused ages were excluded from analysis.
The OFHS methods included measures to ensure representation of all race/ethnicity categories. Ohio’s six largest metropolitan counties were selected for African American oversample. Asian and Hispanic households were also oversampled using a generated surname sample frame.
The results from the survey can be generalized to the population surveyed. All analyses include all surveyed children in Ohio unless otherwise specified. Survey weights were computed for probability of selection, under coverage, and post-stratification adjustments.
The OFHS data enables sub-state analyses which can be useful to policy and decision makers at the community level. This site’s geographic analyses use county identifications based on the sampling frame rather than the respondent’s geocoded address because there were no significant differences between the corrected and uncorrected results. This site includes sub-state analyses examining differences among six metropolitan counties (Cuyahoga, Franklin, Hamilton, Lucas, Montgomery, and Summit); four major regions (Metropolitan, Appalachian, Rural non-Appalachian, and Suburban); and eight managed care regions within Ohio.
Statistical Significance
Measurements compared from 2004 to 2008 are noted to be statistically significant in the text and in tables (measurements in bold) where confidence intervals do not overlap. Intervals for measurements were calculated using SAS version 9.1 at 95% confidence. Therefore, we are 95% certain the true value of the measurement falls within the confidence intervals. Because the surveys in 2004 and 2008 differ slightly in sampling methodology, using the confidence intervals to determine statistical significance is necessary, as opposed to using a statistical test. This method may be a bit more conservative than using a statistical test. Caution should be used when making comparisons between measurements with large confidence intervals, as they tend to have small samples of the population being examined.
Limitations
The OFHS has under coverage in some less affluent counties, causing people with higher incomes to be more widely represented in the data. Post stratification weights have been applied to correct these under coverage issues.
Interpreting OFHS data by specific subgroups should be done with caution. Some subgroup estimates have small numbers for a particular topic and should be used with caution. If the cell size is less than 25, the estimate is not stable and therefore, not provided in this report. Additional checks of the estimates were done to determine if the Relative Standard Error (RSE) was greater than 30%, which is also an indicator that the estimate is unreliable. Because the data is not presented for unstable estimates in this report, there are some subgroup estimates that cannot be compared to previous years or to other subgroups for the same topic.
Data collected with the OFHS were self-reported for adults and parent-reported for children. Information collected on insurance status, health status and behaviors, for example, were not validated with medical records or claims data to ensure accuracy.
Data Gaps
Comparing estimates from 2004 to 2008 is possible where OFHS variables are identical. Most estimates are comparable where they have identical wording and skip patterns across the two survey years. Some variables have minor differences in wording or skip patterns, but are still considered to be identical from one survey to the next. Other indicators have no comparable variables in the previous survey, and therefore, cannot be examined over time. These indicators include:
Overall Mental Health Status of Children and Youth (Health Status)
CSHCN Affected by Current Severe or Moderate Asthma (Health Status)
Body Mass Index for Age Indicates Overweight or Obese (Health Status)
Currently Has Diabetes (Health Status)
Currently Has Asthma (Health Status)
Always or Usually Received Help with Arranging or Coordinating Child's Health Care (Basics of Good Care)
Ever Tested for Lead (Basics of Good Care)
Vision Ever Tested (Basics of Good Care)
Had at Least One Preventive Dental Care Visit in the Past Year (Healthy Development)
Had both Medical AND Dental Preventive Care Visits in the Past 12 Months (Healthy Development)
These instances represent gaps in the data that prohibit readers from determining improvements or declines from one time period to the next. Maintaining a robust survey covering all topics of interest with the same survey design is a key factor for obtaining future estimates that assess improvements or declines.
Further gaps in child health data in Ohio were revealed during the project’s advisory committee meeting, which was held in part to identify priorities in child health and health care in the state. Input was gathered on the top priority indicators for participating stakeholders, which is an important and helpful step in identifying data gaps for this version of the Ohio Chartbook project.
Great interest was expressed in having information on behavioral and emotional health of children in Ohio, including access to providers for this type of need. Further interest was shown for data on Medicaid dollars spent toward mental health services for children. More detailed information on behavioral health of children is available from the National Survey of Children with Special Health Care Needs (NCSHCN) and can be incorporated in the next phase of the Ohio Chartbook. Information on expenditures toward mental health can be obtained from the Medical Expenditure Panel Survey (MEPS), as well as Medicaid claims data. This information can be included in later phases of the project, but will not be available on a sub-state level and may not be available on some sub-group levels, such as by race and ethnicity.
Interest was also expressed in having data to compile a medical home indicator from the OFHS data, which would provide important information regarding access to and quality of care on a state and sub-state level. While, the OFHS does not include questions to compile a medical home indicator, the National Survey of Children’s Health (NSCH) does contain such questions and will be available in late spring of 2009 for the second phase of the Ohio Chartbook. Unfortunately, this data will not be available on a sub-state level.
Having a sense of how well children are being screened in general is also a priority for stakeholders. While the OFHS includes information on lead testing and vision testing, stakeholders said they would benefit from having information on overall health screening of children. The Promoting Healthy Development Survey (PHDS) is a possible source for this information, and can be included in a later phase. However, like other national surveys, data will only be available on a state level.
Stakeholders also stated they would like to see data on substance abuse among children in Ohio. The Youth Risk Behavior Survey (YRBS) can provide some data on this topic for high school students in Ohio, such as alcohol, tobacco, and other drug use. Data from the Ohio YRBS is available at a statewide level only.
Concern was expressed regarding the size of the sample for the OFHS because some populations may not be represented completely, such as Hispanics. Information obtained for this ethnic group is available on a limited basis and may not be examined on smaller levels, such as by most income groups.
While not all priorities can be found in one data source, future researchers may consider the stakeholder priorities mentioned here to help guide decision making in developing resources for gathering information on child health and health care in Ohio.